1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
|
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
; Copyright(c) 2011-2016 Intel Corporation All rights reserved.
;
; Redistribution and use in source and binary forms, with or without
; modification, are permitted provided that the following conditions
; are met:
; * Redistributions of source code must retain the above copyright
; notice, this list of conditions and the following disclaimer.
; * Redistributions in binary form must reproduce the above copyright
; notice, this list of conditions and the following disclaimer in
; the documentation and/or other materials provided with the
; distribution.
; * Neither the name of Intel Corporation nor the names of its
; contributors may be used to endorse or promote products derived
; from this software without specific prior written permission.
;
; THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
; "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
; LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
; A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
; OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
; SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
; LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
; DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
; THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
; (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
; OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
%include "options.asm"
%include "lz0a_const.asm"
%include "data_struct2.asm"
%include "bitbuf2.asm"
%include "huffman.asm"
%include "igzip_compare_types.asm"
%include "reg_sizes.asm"
%include "stdmac.asm"
%ifdef DEBUG
%macro MARK 1
global %1
%1:
%endm
%else
%macro MARK 1
%endm
%endif
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
%define tmp2 rcx
%define hash2 rcx
%define curr_data rax
%define code rax
%define tmp5 rax
%define tmp4 rbx
%define dist rbx
%define code2 rbx
%define hash rdx
%define len rdx
%define code_len3 rdx
%define tmp8 rdx
%define tmp1 rsi
%define code_len2 rsi
%define file_start rdi
%define m_bit_count rbp
%define curr_data2 r8
%define len2 r8
%define tmp6 r8
%define m_bits r9
%define f_i r10
%define m_out_buf r11
%define f_end_i r12
%define dist2 r12
%define tmp7 r12
%define code4 r12
%define tmp3 r13
%define code3 r13
%define stream r14
%define hufftables r15
;; GPR r8 & r15 can be used
%define xtmp0 xmm0 ; tmp
%define xtmp1 xmm1 ; tmp
%define xhash xmm2
%define xmask xmm3
%define xdata xmm4
%define ytmp0 ymm0 ; tmp
%define ytmp1 ymm1 ; tmp
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
blen_mem_offset equ 0 ; local variable (8 bytes)
f_end_i_mem_offset equ 8
gpr_save_mem_offset equ 16 ; gpr save area (8*8 bytes)
xmm_save_mem_offset equ 16 + 8*8 ; xmm save area (4*16 bytes) (16 byte aligned)
stack_size equ 2*8 + 8*8 + 4*16 + 8
;;; 8 because stack address is odd multiple of 8 after a function call and
;;; we want it aligned to 16 bytes
; void isal_deflate_body ( isal_zstream *stream )
; arg 1: rcx: addr of stream
global isal_deflate_body_ %+ ARCH
isal_deflate_body_ %+ ARCH %+ :
%ifidn __OUTPUT_FORMAT__, elf64
mov rcx, rdi
%endif
;; do nothing if (avail_in == 0)
cmp dword [rcx + _avail_in], 0
jne skip1
;; Set stream's next state
mov rdx, ZSTATE_FLUSH_READ_BUFFER
mov rax, ZSTATE_BODY
cmp dword [rcx + _end_of_stream], 0
cmovne rax, rdx
cmp dword [rcx + _flush], _NO_FLUSH
cmovne rax, rdx
mov dword [rcx + _internal_state_state], eax
ret
skip1:
%ifdef ALIGN_STACK
push rbp
mov rbp, rsp
sub rsp, stack_size
and rsp, ~15
%else
sub rsp, stack_size
%endif
mov [rsp + gpr_save_mem_offset + 0*8], rbx
mov [rsp + gpr_save_mem_offset + 1*8], rsi
mov [rsp + gpr_save_mem_offset + 2*8], rdi
mov [rsp + gpr_save_mem_offset + 3*8], rbp
mov [rsp + gpr_save_mem_offset + 4*8], r12
mov [rsp + gpr_save_mem_offset + 5*8], r13
mov [rsp + gpr_save_mem_offset + 6*8], r14
mov [rsp + gpr_save_mem_offset + 7*8], r15
mov stream, rcx
mov dword [stream + _internal_state_has_eob], 0
MOVDQU xmask, [mask]
; state->bitbuf.set_buf(stream->next_out, stream->avail_out);
mov m_out_buf, [stream + _next_out]
mov [stream + _internal_state_bitbuf_m_out_start], m_out_buf
mov tmp1 %+ d, [stream + _avail_out]
add tmp1, m_out_buf
sub tmp1, SLOP
mov [stream + _internal_state_bitbuf_m_out_end], tmp1
mov m_bits, [stream + _internal_state_bitbuf_m_bits]
mov m_bit_count %+ d, [stream + _internal_state_bitbuf_m_bit_count]
mov hufftables, [stream + _hufftables]
mov file_start, [stream + _next_in]
mov f_i %+ d, dword [stream + _total_in]
sub file_start, f_i
mov f_end_i %+ d, [stream + _avail_in]
add f_end_i, f_i
; f_end_i -= LA;
sub f_end_i, LA
mov [rsp + f_end_i_mem_offset], f_end_i
; if (f_end_i <= 0) continue;
cmp f_end_i, f_i
jle input_end
; for (f_i = f_start_i; f_i < f_end_i; f_i++) {
MARK __body_compute_hash_ %+ ARCH
MOVDQU xdata, [file_start + f_i]
mov curr_data, [file_start + f_i]
mov tmp3, curr_data
mov tmp6, curr_data
compute_hash hash, curr_data
shr tmp3, 8
compute_hash hash2, tmp3
and hash, HASH_MASK
and hash2, HASH_MASK
cmp dword [stream + _internal_state_has_hist], 0
je write_first_byte
jmp loop2
align 16
loop2:
; if (state->bitbuf.is_full()) {
cmp m_out_buf, [stream + _internal_state_bitbuf_m_out_end]
ja output_end
xor dist, dist
xor dist2, dist2
xor tmp3, tmp3
lea tmp1, [file_start + f_i]
mov dist %+ w, f_i %+ w
dec dist
sub dist %+ w, word [stream + _internal_state_head + 2 * hash]
mov [stream + _internal_state_head + 2 * hash], f_i %+ w
inc f_i
MOVQ tmp6, xdata
shr tmp5, 16
mov tmp8, tmp5
compute_hash tmp6, tmp5
mov dist2 %+ w, f_i %+ w
dec dist2
sub dist2 %+ w, word [stream + _internal_state_head + 2 * hash2]
mov [stream + _internal_state_head + 2 * hash2], f_i %+ w
; if ((dist-1) < (D-1)) {
and dist %+ d, (D-1)
neg dist
shr tmp8, 8
compute_hash tmp2, tmp8
and dist2 %+ d, (D-1)
neg dist2
MARK __body_compare_ %+ ARCH
;; Check for long len/dist match (>7) with first literal
MOVQ len, xdata
mov curr_data, len
PSRLDQ xdata, 1
xor len, [tmp1 + dist - 1]
jz compare_loop
MOVD xhash, tmp6 %+ d
PINSRD xhash, tmp2 %+ d, 1
PAND xhash, xhash, xmask
;; Check for len/dist match (>7) with second literal
MOVQ len2, xdata
xor len2, [tmp1 + dist2]
jz compare_loop2
;; Specutively load the code for the first literal
movzx tmp1, curr_data %+ b
get_lit_code tmp1, code3, rcx, hufftables
;; Check for len/dist match for first literal
test len %+ d, 0xFFFFFFFF
jz len_dist_huffman_pre
;; Specutively load the code for the second literal
shr curr_data, 8
and curr_data, 0xff
get_lit_code curr_data, code2, code_len2, hufftables
SHLX code2, code2, rcx
or code2, code3
add code_len2, rcx
;; Check for len/dist match for second literal
test len2 %+ d, 0xFFFFFFFF
jnz write_lit_bits
MARK __body_len_dist_lit_huffman_ %+ ARCH
len_dist_lit_huffman_pre:
mov code_len3, rcx
bsf len2, len2
shr len2, 3
len_dist_lit_huffman:
neg dist2
%ifndef LONGER_HUFFTABLE
mov tmp4, dist2
get_dist_code tmp4, code4, code_len2, hufftables ;; clobbers dist, rcx
%else
get_dist_code dist2, code4, code_len2, hufftables
%endif
get_len_code len2, code, rcx, hufftables ;; rcx is code_len
SHLX code4, code4, rcx
or code4, code
add code_len2, rcx
add f_i, len2
neg len2
MOVQ tmp5, xdata
shr tmp5, 24
compute_hash tmp4, tmp5
and tmp4, HASH_MASK
SHLX code4, code4, code_len3
or code4, code3
add code_len2, code_len3
;; Setup for updating hash
lea tmp3, [f_i + len2 + 1] ; tmp3 <= k
MOVDQU xdata, [file_start + f_i]
mov curr_data, [file_start + f_i]
mov curr_data2, curr_data
MOVD hash %+ d, xhash
PEXTRD hash2 %+ d, xhash, 1
mov [stream + _internal_state_head + 2 * hash], tmp3 %+ w
compute_hash hash, curr_data
add tmp3,1
mov [stream + _internal_state_head + 2 * hash2], tmp3 %+ w
add tmp3, 1
mov [stream + _internal_state_head + 2 * tmp4], tmp3 %+ w
write_bits m_bits, m_bit_count, code4, code_len2, m_out_buf, tmp4
mov f_end_i, [rsp + f_end_i_mem_offset]
shr curr_data2, 8
compute_hash hash2, curr_data2
%ifdef NO_LIMIT_HASH_UPDATE
loop3:
add tmp3,1
cmp tmp3, f_i
jae loop3_done
mov tmp6, [file_start + tmp3]
compute_hash tmp4, tmp6
and tmp4 %+ d, HASH_MASK
; state->head[hash] = k;
mov [stream + _internal_state_head + 2 * tmp4], tmp3 %+ w
jmp loop3
loop3_done:
%endif
; hash = compute_hash(state->file_start + f_i) & HASH_MASK;
and hash %+ d, HASH_MASK
and hash2 %+ d, HASH_MASK
; continue
cmp f_i, f_end_i
jl loop2
jmp input_end
;; encode as dist/len
MARK __body_len_dist_huffman_ %+ ARCH
len_dist_huffman_pre:
bsf len, len
shr len, 3
len_dist_huffman:
dec f_i
neg dist
; get_dist_code(dist, &code2, &code_len2);
%ifndef LONGER_HUFFTABLE
mov tmp3, dist ; since code2 and dist are rbx
get_dist_code tmp3, code2, code_len2, hufftables ;; clobbers dist, rcx
%else
get_dist_code dist, code2, code_len2, hufftables
%endif
; get_len_code(len, &code, &code_len);
get_len_code len, code, rcx, hufftables ;; rcx is code_len
; code2 <<= code_len
; code2 |= code
; code_len2 += code_len
SHLX code2, code2, rcx
or code2, code
add code_len2, rcx
;; Setup for updateing hash
lea tmp3, [f_i + 2] ; tmp3 <= k
add f_i, len
MOVD hash %+ d, xhash
PEXTRD hash2 %+ d, xhash, 1
mov [stream + _internal_state_head + 2 * hash], tmp3 %+ w
add tmp3,1
mov [stream + _internal_state_head + 2 * hash2], tmp3 %+ w
MOVDQU xdata, [file_start + f_i]
mov curr_data, [file_start + f_i]
mov curr_data2, curr_data
compute_hash hash, curr_data
write_bits m_bits, m_bit_count, code2, code_len2, m_out_buf, tmp7
mov f_end_i, [rsp + f_end_i_mem_offset]
shr curr_data2, 8
compute_hash hash2, curr_data2
%ifdef NO_LIMIT_HASH_UPDATE
loop4:
add tmp3,1
cmp tmp3, f_i
jae loop4_done
mov tmp6, [file_start + tmp3]
compute_hash tmp4, tmp6
and tmp4, HASH_MASK
mov [stream + _internal_state_head + 2 * tmp4], tmp3 %+ w
jmp loop4
loop4_done:
%endif
; hash = compute_hash(state->file_start + f_i) & HASH_MASK;
and hash %+ d, HASH_MASK
and hash2 %+ d, HASH_MASK
; continue
cmp f_i, f_end_i
jl loop2
jmp input_end
MARK __body_write_lit_bits_ %+ ARCH
write_lit_bits:
MOVDQU xdata, [file_start + f_i + 1]
mov f_end_i, [rsp + f_end_i_mem_offset]
add f_i, 1
mov curr_data, [file_start + f_i]
MOVD hash %+ d, xhash
write_bits m_bits, m_bit_count, code2, code_len2, m_out_buf, tmp3
PEXTRD hash2 %+ d, xhash, 1
; continue
cmp f_i, f_end_i
jl loop2
input_end:
mov tmp1, ZSTATE_FLUSH_READ_BUFFER
mov tmp5, ZSTATE_BODY
cmp dword [stream + _end_of_stream], 0
cmovne tmp5, tmp1
cmp dword [stream + _flush], _NO_FLUSH
cmovne tmp5, tmp1
mov dword [stream + _internal_state_state], tmp5 %+ d
output_end:
;; update input buffer
add f_end_i, LA
mov [stream + _total_in], f_i %+ d
add file_start, f_i
mov [stream + _next_in], file_start
sub f_end_i, f_i
mov [stream + _avail_in], f_end_i %+ d
;; update output buffer
mov [stream + _next_out], m_out_buf
sub m_out_buf, [stream + _internal_state_bitbuf_m_out_start]
sub [stream + _avail_out], m_out_buf %+ d
add [stream + _total_out], m_out_buf %+ d
mov [stream + _internal_state_bitbuf_m_bits], m_bits
mov [stream + _internal_state_bitbuf_m_bit_count], m_bit_count %+ d
mov rbx, [rsp + gpr_save_mem_offset + 0*8]
mov rsi, [rsp + gpr_save_mem_offset + 1*8]
mov rdi, [rsp + gpr_save_mem_offset + 2*8]
mov rbp, [rsp + gpr_save_mem_offset + 3*8]
mov r12, [rsp + gpr_save_mem_offset + 4*8]
mov r13, [rsp + gpr_save_mem_offset + 5*8]
mov r14, [rsp + gpr_save_mem_offset + 6*8]
mov r15, [rsp + gpr_save_mem_offset + 7*8]
%ifndef ALIGN_STACK
add rsp, stack_size
%else
mov rsp, rbp
pop rbp
%endif
ret
MARK __body_compare_loops_ %+ ARCH
compare_loop:
MOVD xhash, tmp6 %+ d
PINSRD xhash, tmp2 %+ d, 1
PAND xhash, xhash, xmask
lea tmp2, [tmp1 + dist - 1]
%if (COMPARE_TYPE == 1)
compare250 tmp1, tmp2, len, tmp3
%elif (COMPARE_TYPE == 2)
compare250_x tmp1, tmp2, len, tmp3, xtmp0, xtmp1
%elif (COMPARE_TYPE == 3)
compare250_y tmp1, tmp2, len, tmp3, ytmp0, ytmp1
%else
%error Unknown Compare type COMPARE_TYPE
% error
%endif
jmp len_dist_huffman
compare_loop2:
lea tmp2, [tmp1 + dist2]
add tmp1, 1
%if (COMPARE_TYPE == 1)
compare250 tmp1, tmp2, len2, tmp3
%elif (COMPARE_TYPE == 2)
compare250_x tmp1, tmp2, len2, tmp3, xtmp0, xtmp1
%elif (COMPARE_TYPE == 3)
compare250_y tmp1, tmp2, len2, tmp3, ytmp0, ytmp1
%else
%error Unknown Compare type COMPARE_TYPE
% error
%endif
and curr_data, 0xff
get_lit_code curr_data, code3, code_len3, hufftables
jmp len_dist_lit_huffman
MARK __write_first_byte_ %+ ARCH
write_first_byte:
cmp m_out_buf, [stream + _internal_state_bitbuf_m_out_end]
ja output_end
mov dword [stream + _internal_state_has_hist], 1
mov [stream + _internal_state_head + 2 * hash], f_i %+ w
mov hash, hash2
shr tmp6, 16
compute_hash hash2, tmp6
MOVD xhash, hash %+ d
PINSRD xhash, hash2 %+ d, 1
PAND xhash, xhash, xmask
and curr_data, 0xff
get_lit_code curr_data, code2, code_len2, hufftables
jmp write_lit_bits
section .data
align 16
mask: dd HASH_MASK, HASH_MASK, HASH_MASK, HASH_MASK
const_D: dq D
|